Department of Electrical Engineering and Computer Science



Dr. Richard Mills


Background

I am currently a computational scientist in the Computational Earth Sciences Group of the Computer Science and Mathematics Division, and the Earth and Aquatic Sciences Group of the Environmental Sciences Division at Oak Ridge National Laboratory. I came to ORNL in 2004 after completing my Ph.D. in computer science (with a specialization in computational science) under the direction of Andreas Stathopoulos at the College of William and Mary in beautiful Williamsburg, Virginia. My graduate studies were supported by a Department of Energy Computational Science Graduate Fellowship administered by the fine folks at Krell Institute, and for this I shall be forever grateful. I was fortunate enough to complete my fellowship practicum at Los Alamos National Laboratory under the masterful tutelage of Peter Lichtner of the Hydrology, Geochemistry and Geology group. Prior to my graduate studies, I studied geology and physics (in addition to Volunteers Football, of course) in the hospitable environs of the University of Tennessee, Knoxville, as a Chancellor's Scholar.

My broad interest is in developing and applying numerical methods and software to enable the solution of very computationally challenging problems that arise in the natural sciences. My specific interests as a computer scientist arise from this broad interest and mostly fall into two categories. The first is numerical methods and software for the solution of partial differential equations, and particularly in iterative methods for solving the sparse algebraic systems that arise from their discretization. The second is parallel implementation issues and the interplay between numerical algorithms and computer architecture. Algorithms designed with one kind of architecture in mind may perform very poorly on another. For instance, codes that are very efficient on superscalar processors might perform extremely poorly on vector processor machines. My interest is in developing scalable, parallel numerical applications that not only possess good numerical properties, but are also tailored to effectively utilize their target resources.



 

The University of Tennessee, Knoxville. Big Orange. Big Ideas.

Knoxville, Tennessee 37996 | 865-974-1000
The flagship campus of the University of Tennessee System